Bridging the Gap Between Model Assumptions and Realities in Leak Localization for Water Networks
Round 1
Reviewer 1 Report
Comments and Suggestions for AuthorsI congratulate the authors on the idea for this article. The topic they addressed is very actual and still unresolved. Like many other researchers, I also encountered the problem of unrealistic assumptions and sensor noise. The article will certainly be of interest to both researchers and water utilities. I read it with great interest and pleasure.
The introduction provides a good introduction to the article. The purpose of the article is clearly stated.
The materials and methods were adopted correctly by the authors. A reviewer would have adopted a similar methodology.
- Simplifying assumptions in leak location methods
The authors focus here on stationary or EPS models, as they are the most common in water supply network management. However, there is another class of models – dynamic models, which reflect transient states and can exhibit momentary pressure spikes. It is worth addressing this class of models in the future.
The reviewer encountered a situation in which even a large leak (equivalent to opening a fire hydrant) did not cause pressure gauge readings to change beyond its tolerance. We are not talking about several meters here, but rather a single centimeters of pressure height. This was due to the encountered geometric structure of the network and its concomitant oversizing. As a result, the manometer had to be abandoned as the primary diagnostic device in favor of acoustic correlators.
Tables 1-5 are not cited in the text.
4 Conclusions
The conclusions are clear and well documented in the text.
Author Response
Reviewer 1
I congratulate the authors on the idea for this article. The topic they addressed is very actual and still unresolved. Like many other researchers, I also encountered the problem of unrealistic assumptions and sensor noise. The article will certainly be of interest to both researchers and water utilities. I read it with great interest and pleasure.
The introduction provides a good introduction to the article. The purpose of the article is clearly stated.
The materials and methods were adopted correctly by the authors. A reviewer would have adopted a similar methodology.
Response: Thanks to the reviewer for the comment.
- Simplifying assumptions in leak location methods
The authors focus here on stationary or EPS models, as they are the most common in water supply network management. However, there is another class of models – dynamic models, which reflect transient states and can exhibit momentary pressure spikes. It is worth addressing this class of models in the future.
The reviewer encountered a situation in which even a large leak (equivalent to opening a fire hydrant) did not cause pressure gauge readings to change beyond its tolerance. We are not talking about several meters here, but rather a single centimeters of pressure height. This was due to the encountered geometric structure of the network and its concomitant oversizing. As a result, the manometer had to be abandoned as the primary diagnostic device in favor of acoustic correlators.
Response: Thanks to the reviewer for the comment. We have underlined this aspect in the manuscript at lines 386-400.
Tables 1-5 are not cited in the text.
Response: Thanks to the reviewer for the comment. All tables are now cited in the text.
4 Conclusions
The conclusions are clear and well documented in the text.
Response: Thanks to the reviewer for the comment.
Reviewer 2 Report
Comments and Suggestions for AuthorsThe manuscript is relevant and the type of manuscript is appropriate (review). My major comments are:
- The paper must include a revision of the importance of digital twin as an important tool for minimizing and analyzing water leakages.
- It is of utmost important to present the difference between leaks than depends on pressure and neglecting it.
- The paper must present a diagram that presents the methodology used in this research. I hope a complex diagram that shows in details this research.
- The manuscript must present some figures around the world that contains many citations presented in this research.
- The English must be revised (i.e, line 37 there is a typo (“Minimizing” instead of “minimizing”).
- The paper must be indicates the limitations of current models that only considers leakages using Extended Period Simulations (EPS). The effects of regulating valves must be presented.
Author Response
Reviewer 2
The manuscript is relevant and the type of manuscript is appropriate (review). My major comments are:
- The paper must include a revision of the importance of digital twin as an important tool for minimizing and analyzing water leakages.
Response: Thanks to the reviewer for the comment. We have underlined this aspect in the manuscript at lines 510-520.
- It is of utmost important to present the difference between leaks than depends on pressure and neglecting it.
Response: Thanks to the reviewer for the comment. We have underlined this aspect in the manuscript at lines 386-400.
- The paper must present a diagram that presents the methodology used in this research. I hope a complex diagram that shows in details this research.
Response: Thanks to the reviewer for the comment. The flowchart has been added to line 150.
- The manuscript must present some figures around the world that contains many citations presented in this research.
Response: Thanks to the reviewer for the comment. We have underlined this aspect in the manuscript at lines 38-40.
- The English must be revised (i.e, line 37 there is a typo (“Minimizing” instead of “minimizing”).
Response: Thanks to the reviewer for the comment. The manuscript has been revised.
- The paper must be indicates the limitations of current models that only considers leakages using Extended Period Simulations (EPS). The effects of regulating valves must be presented.
Response: Thanks to the reviewer for the comment. We have underlined this aspect in the manuscript at lines 386-400.
Reviewer 3 Report
Comments and Suggestions for AuthorsThe manuscript presents a critical review of leak localisation methods in pressurised water distribution networks.
The paper discusses “advanced approaches” (hybrid, model-informed ML, PINNs, benchmarking) that attempt to relax these assumptions and better reflect real operational conditions. This angle—explicitly structuring the review around model assumptions vs. field realities—is interesting and potentially valuable for practitioners and researchers working on real-world leak management.However, in its current form the manuscript needs clearer positioning relative to existing reviews, a more rigorous and transparent description of the review methodology. These issues, in my view, justify a recommendation of major revision.
Tables 1–5 provide a useful summary of recent works. However, for ease of cross-referencing, please ensure that every work mentioned in the text of these tables is accompanied by its corresponding citation number in the table, so that readers can readily locate the full references in the bibliography.
Section 2 (“Materials and Methods”) currently describes the process only in general terms.Even for a narrative review, readers will expect a clearer and more reproducible methodology: Specify databases and time window more explicitly.Consider adding a simple flow diagram (not necessarily fully PRISMA-compliant, but in that spirit) showing the screening process: number of records identified, screened, excluded, and finally retained.
Section 3 is nicely structured into four key assumptions, each with sub-discussion and a summary table (Tables 1–4), which is a strong point of the paper. However, some improvements could make the structure clearer:
In practice, uncertain demands, noisy measurements, multiple leaks, and temporal dynamics often interact. You partially note this, but it may be worth adding a short bridging subsection (e.g., 3.6 “Interplay among assumptions”) highlighting questions such as how demand uncertainty and sensor noise are often indistinguishable in pressure residues,how multiple leaks + temporal evolution complicate detection thresholds and change-detection methods.
Some references and arguments reappear in multiple places. Try to centralise cross-cutting points and avoid restating the same idea with similar wording in several subsections.
The “Advanced approaches beyond traditional standards” (Section 3.5) is very rich but somewhat dense. Consider explicitly linking each advanced approach back to which assumptions it mitigates (e.g., PINNs tackle demand uncertainty and imperfect models; hybrid schemes address sparse data and incomplete models, etc.).
Author Response
The manuscript presents a critical review of leak localisation methods in pressurised water distribution networks.
The paper discusses “advanced approaches” (hybrid, model-informed ML, PINNs, benchmarking) that attempt to relax these assumptions and better reflect real operational conditions. This angle—explicitly structuring the review around model assumptions vs. field realities—is interesting and potentially valuable for practitioners and researchers working on real-world leak management.However, in its current form the manuscript needs clearer positioning relative to existing reviews, a more rigorous and transparent description of the review methodology. These issues, in my view, justify a recommendation of major revision.
Tables 1–5 provide a useful summary of recent works. However, for ease of cross-referencing, please ensure that every work mentioned in the text of these tables is accompanied by its corresponding citation number in the table, so that readers can readily locate the full references in the bibliography.
Response: Thanks to the reviewer for the comment. The citations have been inserted into the tables 1–5 .
Section 2 (“Materials and Methods”) currently describes the process only in general terms.Even for a narrative review, readers will expect a clearer and more reproducible methodology: Specify databases and time window more explicitly.Consider adding a simple flow diagram (not necessarily fully PRISMA-compliant, but in that spirit) showing the screening process: number of records identified, screened, excluded, and finally retained.
Response: Thanks to the reviewer for the comment. The flowchart has been added to line 150.
Section 3 is nicely structured into four key assumptions, each with sub-discussion and a summary table (Tables 1–4), which is a strong point of the paper. However, some improvements could make the structure clearer:
In practice, uncertain demands, noisy measurements, multiple leaks, and temporal dynamics often interact. You partially note this, but it may be worth adding a short bridging subsection (e.g., 3.6 “Interplay among assumptions”) highlighting questions such as how demand uncertainty and sensor noise are often indistinguishable in pressure residues,how multiple leaks + temporal evolution complicate detection thresholds and change-detection methods.
Response: Thanks to the reviewer for the comment. The new section has been added at line 642.
Some references and arguments reappear in multiple places. Try to centralise cross-cutting points and avoid restating the same idea with similar wording in several subsections.
Response: Thanks to the reviewer for the comment. We have tried to centralize the cross-cutting points at line 470.
The “Advanced approaches beyond traditional standards” (Section 3.5) is very rich but somewhat dense. Consider explicitly linking each advanced approach back to which assumptions it mitigates (e.g., PINNs tackle demand uncertainty and imperfect models; hybrid schemes address sparse data and incomplete models, etc.).
Response: Thanks to the reviewer for the comment. Table 5 has been added to explicitly link each advanced approach to the assumptions it relaxes.
Round 2
Reviewer 2 Report
Comments and Suggestions for AuthorsDear authors,
You must improve the complexity of Figure 1. Include all relationships of the steps of your methodology. Considers different decisions in the diagram (Yes and No). In addition, improve the quality of the presentation,
Author Response
You must improve the complexity of Figure 1. Include all relationships of the steps of your methodology. Considers different decisions in the diagram (Yes and No). In addition, improve the quality of the presentation,
Response: Thanks to the reviewer for the comments. Figure 1 has been improved. Additionally, Section 2 has been revised at lines 128-181 to improve the quality of the presentation.
Reviewer 3 Report
Comments and Suggestions for AuthorsThe author has addressed my questions very well, and I believe the manuscript can be accepted.
Author Response
The author has addressed my questions very well, and I believe the manuscript can be accepted.
Response: The authors thank the reviewer for his valuable comments.
